Inline optimization options change output

[kirbychapman]

I have a program that i have been running in debug mode. I just changed it to release mode and many of the numbers changed to unrealistic values (this is an iterative solution code, so numbers blow up pretty quickly). I narrowed it down to the inline optimization in the release version. It defaults to "any suitable." This causes the problem. If I change it to "Only INLINE Directive (/Ob1)" then everything is fine.

So while the release version now works, i would still like to understand the differences. There are numerous subroutines and functions in the code, if that is meaningful.

Thanks for any guidance.

[Steven_L_Intel1]

Without seeing the actual program, it's impossible to know. Inline expansion should not change the results of a correct program. Your program may be incorrect, or there may be a compiler bug. Which specific compiler version are you using?

[kirbychapman]

Understandable that it is difficult to diagnose without the code. I am using 11.1.051.

Is there a way that i can designate which functions are treated in a specific inlining way? That wasy i may be able to narrow it down to the problem child. Alternatively, i can send you the source code and sln file...

Thanks for the help, Steve -- i'm beginning to think you live online!

[Steven_L_Intel1]

Please try 11.1.065 and see if the problem remains. I am aware of an inlining bug fixed in that version.